Alchemy: Bounded Enumeration and Interactive Exploration of Small-Molecule Radical Reaction Networks
Abstract Automated construction of chemical reaction networks is limited by the combinatorial growth of possible molecules, intermediates, and formal transformations. We present alchemy, an open-source Python workflow for systematic enumeration and interactive exploration of small-molecule radical reaction networks. Within an explicitly bounded graph space, alchemy generates connected, neutral C/N/H molecular graphs with single, double, or triple bonds and specified radical counts; RDKit sanitization and canonicalization then remove graph-invalid and duplicate records. Formal reaction candidates are constructed without learned reaction templates by combining bond-dissociation records into rearrangement and transfer stoichiometries. Parallel generation, persistent disk-backed storage, and Kuzu graph representation make the resulting formal network searchable. An independent brute-force control recovers exactly the same 416 neutral molecular graphs as alchemy in a three-heavy-atom C/N/H space. In a trimethylamine case study limited to at most five heavy atoms and one unpaired electron per reaction side, the workflow generated 41,735 unique molecular graphs and a seed-0 bounded sample of 60,035 first-generation reaction records. The reaction construction is not complete because high-frequency shared fragments are handled by a documented deterministic sampling cap. An optional PaiNN molecular-energy model is distributed with the software, but its closed-shell QM9 benchmark does not validate the open-shell species considered here, and its output is not used to rank, filter, or interpret the case-study records. Alchemy therefore, provides a reproducible route to bounded reaction-space construction and downstream graph inspection.
Authors
- Mega Kar (ORCID: https://orcid.org/0000-0003-4927-6254)
- Sherif Abdulkader Tawfik (ORCID: https://orcid.org/0000-0003-3592-1419)
- Truyen Tran (ORCID: https://orcid.org/0000-0001-6531-8907)
- Le Minh Vu
- Priya Sharmesh
Institutions
- Deakin University (AU)
- Phenikaa University (VN)
- VinUniversity (VN)
- Monash University (AU)
Publication Details
- Journal
- Journal of Chemical Information and Modeling
- Published
- 2026-11-13
- DOI
- https://doi.org/10.1021/acs.jcim.6c02462
- Primary Topic
- Computational Drug Discovery Methods
- Type
- article
- Field-Weighted Citation Impact
- 0.00